go.scatterpolar : trying to render radar graph with various lines color not working

Question

I am trying to build a radar chart where each line is of different color.

I feel like I have followed the doc closely and I am now facing an error I can't seem solve, especially because NO ERROR is output!

here is some dummy data I am working with :

r = [52,36,85]
theta = ["analytique", "analogique", "affectif"]

colors = ["blue", "red","yellow"]

Here is what I have for my graph:

for i in range(len(theta)):
    fig_reception.add_trace(go.Scatterpolar(
        mode='lines+text',
        theta=[theta[i]],
        r=[r[i]],
     ,   line_color=text_colors[i],
        fillcolor='#d3d3d3',
        marker=dict(color=text_colors),
    ))


fig_reception.update_layout(autosize=False,
                            height=305,
                  polar=dict(radialaxis = dict(range=[0,100],visible = False),
                                          angularaxis=dict(rotation=180,direction="clockwise") )
                            )
fig_reception.update_layout(
    template=None,
    polar = dict(bgcolor = "rgba(255, 255, 255, 0.2)"),)

fig_reception.update_layout(
    font=dict(
        size=16,
        color="black",
        family="Courier New, monospace",

    ),
    title="Réception",
    title_font_family="Courier New, monospace",
    showlegend=False
)

what's strange its that when I hover each line, a frame with the right color and value shows up.

Here is a picture

Answer 1

I don't have a full solution for you, but I hope my answer leads you in the right way.

Simple start

First, let's simplify and plot a radar/spyder plot with default colors:

import plotly.express as px
import pandas as pd

r = [52,36,85]
theta = ["analytique", "analogique", "affectif"]
types = ["one", "two","three"]
df = pd.DataFrame(dict(r=r, theta=theta, type=types))
df

	r	theta	type
0	52	analytique	one
1	36	analogique	two
2	85	affectif	three

Plotting this with plotly.express.line_polar, gives:

fig = px.line_polar(df, r='r', theta='theta', line_close=True, markers=True)
fig.show()

Every edge its own color

Now, you want every edge to have it's own color. For the sake of this example, I assume you want this color to be based on the column type which I defined earlier.

Simply plotting this straight away will not work, it will only give you the dots, no lines:

fig = px.line_polar(df, r='r', theta='theta', line_close=True, color='type', markers=True)
fig.show()

You need to duplicate the rows, and assign sequential data points the same type.

# First append the df to itself, but only keep the r and theta columns
# This will make the type column NaN for the appended rows
df2 = pd.concat([df, df[['r', 'theta']]]).sort_values(by=['r', 'theta'])
# Now fill the NaN type value by taking the type value of the next row
df2.type.fillna(method='bfill', inplace=True)
# The last type value should be equal to the first type value to close the loop
# This needs to be set manually
df2.type.fillna(df2.type.iloc[0], inplace=True)
df2

	r	theta	type
1	36	analogique	two
1	36	analogique	one
0	52	analytique	one
0	52	analytique	three
2	85	affectif	three
2	85	affectif	two

Now if you plot that, you will get a triangle with every edge having a separate color:

fig = px.line_polar(df2, r='r', theta='theta', color='type', line_close=True, markers=True)
fig.show()

Not sure why the categories have changed order, but you can probably fix that by sorting the df2 DataFrame differently.

Text labels

If you would like to have text labels in your graph, you'll find in the docs that there is a text parameter:

fig = px.line_polar(df2, r='r', theta='theta', color='type', text='r', line_close=True, markers=True)
fig.update_traces(textposition='top center')